Update app.py

app.py

@@ -38,7 +38,7 @@ On 🌐Github: [Tonic-AI](https://github.com/tonic-ai) & contribute to🌟 [Buil
 """
 
 # Initialize model and tokenizer
-model_id = "OpenLLM-France/Lucie-7B-Instruct"
+model_id = "OpenLLM-France/Lucie-7B-Instruct-v1.1"
 device = "cuda" if torch.cuda.is_available() else "cpu"
 
 # Get the token from environment variables
@@ -77,13 +77,29 @@ def format_model_info(config):
                 value = value.name
             info.append(f"**{key}:** {value}")
     return "\n".join(info)
+
     
+
+<s><|start_header_id|>system<|end_header_id|>
+
+{SYSTEM}<|eot_id|><|start_header_id|>user<|end_header_id|>
+
+{INPUT}<|eot_id|><|start_header_id|>assistant<|end_header_id|>
+
+{OUTPUT}<|eot_id|>
+
 @spaces.GPU
 def generate_response(system_prompt, user_prompt, temperature, max_new_tokens, top_p, repetition_penalty, top_k):
     # Construct the full prompt with system and user messages
-    full_prompt = f"""<|system|>{system_prompt}</s>
-<|user|>{user_prompt}</s>
-<|assistant|>"""
+    full_prompt = f"""<s><|start_header_id|>system<|end_header_id|>
+
+{user_prompt}<|eot_id|><|start_header_id|>user<|end_header_id|>
+
+{user_prompt}<|eot_id|><|start_header_id|>assistant<|end_header_id|>"""
+
+# <|system|>{system_prompt}</s>
+# <|user|>{user_prompt}</s>
+# <|assistant|>"""
     
     # Prepare the input prompt
     inputs = tokenizer(full_prompt, return_tensors="pt").to(device)